Image treatment process to obtain the Kinematic Parameters During Human Locomotion.

Radu, Ciprian ; Rosca, Ileana

1. INTRODUCTION

The systematic construction of biomechanical models of bipedal locomotion and corresponding stability criteria in the sense of gait synthesis took place in the 1970s in the works of Stepanjenko and Vukobratonic. The biomechanical model consists of few segments, the lower extremities being represented quite faithfully (the upper leg, the lower leg and the foot), while the total upper body torso is be represented by just one segment (Medved, 2001).

Our solution to construct a biomechanical model of lower extremities consists in representing the upper leg and lower leg as a straight beam and the foot as a triangle (the ankle joint, the heel and second metatarsal). Our model is simple and is very useful to determine the most important kinematic parameters during normal locomotion, such as: the angles between anatomical segments, respectively the angle between foot and ground (Rosca & Radu, 2008).

2. METHODOLOGY AND RESULTS

In order to determine the position of a particular observed point experimentally, a stereometric procedure has to be performed. With the help of 3D positions of a one point on the rigid body, the position of this body is determinated. In applying this principle, markers are attached to the subject's body and the positions of the characteristic body landmarks are extracted later from recordings obtained by using a high-speed video camera (Panagiotakis & Tziritas, 2004). Video recording of the light markers attached to the lower right limb (hip, knee and ankle joints heel and the second metatarsal's pick), of a human subject during locomotion, is possible by using a high speed video camera, in special conditions (1/250 frames/second and an adjustable depth recording field of 3500 mm). The recorded images are transmitted to the acquisition system and saved as .avi format, by CamLink software, which is the interface software for high-speed video camera.

For our experiment, the movie length is 8.24 seconds that means 247-recorded frames (for a displayed rate of 30 frames/second in normal conditions). The experimental solution to record the human locomotion was presented in the first scientific paper, called: On the Design of an Experimental Installation to Determine the Kinematic Parameters During Human Locomotion, published in the same proceeding.

The recorded images of human locomotion event in .avi format are imported in Adobe After Effects 6.5 software for image treatment technique, to determine some kinematic parameters.

The most important module of editing software is tracking module. This module has been used by as to track the light markers from the right foot of human subject. In this way, for each light marker we have generated a virtual tracking marker (track1, track2 ... track5) (see figure 1). This virtual tracking marker is positioned in light marker's centre. For an accurate measure, both centers of light marker and virtual tracking marker have to coincide (James, 2001).

Once they are generated and virtual tracking markers have been edited, the next step is to establishing the tracking mode. For our case we have used the color-tracking mode. In this way, we have established that the white color of the light markers to be tracked by virtual tracking markers, during all length movie.

[FIGURE 1 OMITTED]

As it can be seen in figure 1, the tracking markers will follow the light markers, finally obtaining the trajectories of the most important landmarks of lower right limb. Also, as it can be seen in the same figure, there are some major differences regarding the tracking points appearing in the trajectories. The different velocity of the light markers attached to the anatomical parts can explain that.

The velocities of the light markers attached on the foot segment (on the heel, ankle joint and second metatarsal) are higher then velocities of the light markers attached to the knee and hip joint during swing phase of lower limb during locomotion (Radu et al., 2008).Because there are five distinct phases of human normal locomotion, we have determinated for all five of them the angles between anatomical segments, respectively between foot and ground (see figure 3), as follow:

* Phase one (initial contact): [alpha] = 180[degrees], [beta] = 143[degrees], [phi] = 16[degrees], [delta] = 94[degrees];

* Phase two (loading): [alpha] = 166[degrees], [beta] = 138[degrees], [phi] = 0[degrees], [delta] = 95[degrees];

* Phase three (midstance): [alpha] = 169[degrees], [beta] = 144[degrees], [phi] = 0[degrees], [delta] = 91[degrees];

* Phase four (heel off): [alpha] = 166[degrees], [beta] = 132[degrees], [phi] = 10[degrees], [delta] = 87[degrees];

* Phase five (toe off): [alpha] = 161[degrees], [beta] = 120[degrees], [phi] = 24[degrees], [delta] = 87[degrees];

[FIGURE 2 OMITTED]

3. CONCLUSIONS

Our solution to construct the biomechanical model of lower extremity, respectively to obtain some kinematic parameters during human locomotion, is a very simple and accurate method, that because the recorded images are affected by deformation errors in proportion of 0.857%, which is suitable for the error limits admissible for this measuring method.

4. REFERENCES

James, R.G. (2001). Gait Analysis in Cerebral Palsy. Mac Keith Press, ISBN 0 90126 90 8, London, U.K.

Medved, V. (2001). Measurement of Human Locomotion, Congress Library, ISBN 0-8493-7675-0, USA.

Panagiotakis, C. & Tziritas G. (2004). Recognition and Tracking of the Members of a Moving Human Body, Springer Publisher, ISSN 0302-9743, Berlin, Germany.

Radu, M.I, Atanasiu, V., Budescu, E & Ibanescu, R. (2008). Aspects Concerning the Biomechanics of Ankle Joint. Published by Gheorghe Asachi Technical University from Iasi, ISSN 1011-2855, Romania.

Rosca, I.C & Radu, C. (2008). On a Mechatronic System to Determine Dynamic Parameters of the Human Limb During Locomotion Used For Artificial Muscles and Implant Design, 6th International DAAAM Baltic Conference, ISBN 978-9985-59-783-5, Tallin, Estonia.